From e253029ad0bd9a3a7cb4f477c5c8d2d23672188c Mon Sep 17 00:00:00 2001
From: Bruno Oliveira <nicoddemus@gmail.com>
Date: Wed, 15 May 2019 20:08:28 -0300
Subject: [PATCH] Handle lone surrogate unicode character not being
representable in Jython
No tests for this because we don't test with Jython currently.
Fix #5256
---
changelog/5256.bugfix.rst | 1 +
src/_pytest/terminal.py | 10 +++++++++-
2 files changed, 10 insertions(+), 1 deletion(-)
create mode 100644 changelog/5256.bugfix.rst
diff --git a/changelog/5256.bugfix.rst b/changelog/5256.bugfix.rst
new file mode 100644
index 000000000..4df83454e
--- /dev/null
+++ b/changelog/5256.bugfix.rst
@@ -0,0 +1 @@
+Handle internal error due to a lone surrogate unicode character not being representable in Jython.
diff --git a/src/_pytest/terminal.py b/src/_pytest/terminal.py
index 771e6a835..1780792b4 100644
--- a/src/_pytest/terminal.py
+++ b/src/_pytest/terminal.py
@@ -998,7 +998,15 @@ def _get_line_with_reprcrash_message(config, rep, termwidth):
# u'😄' will result in a High Surrogate (U+D83D) character, which is
# rendered as u'�'; in this case we just strip that character out as it
# serves no purpose being rendered
- msg = msg.rstrip(u"\uD83D")
+ try:
+ surrogate = six.unichr(0xD83D)
+ msg = msg.rstrip(surrogate)
+ except ValueError: # pragma: no cover
+ # Jython cannot represent this lone surrogate at all (#5256):
+ # ValueError: unichr() arg is a lone surrogate in range
+ # (0xD800, 0xDFFF) (Jython UTF-16 encoding)
+ # ignore this case as it shouldn't appear in the string anyway
+ pass
msg += ellipsis
line += sep + msg
return line